Talk:Metalloid/Archive 1

Recent edits
Speaking as a non-scientist with just a bit of background in minerology but a strong background in English and writing, I'd like to make some comments on several recent edits.

1) For the non-expert reader, it may not be clear what a "rigorous definition" means. I think "universally agreed upon" clarifies that. Also, that phrase goes along with what the article says, that quite a few scientists have defined metalloid differently.

2) I really don't understand the change from "can be extracted" to "can be obtained". The verb "obtained" has such a general meaning. It can mean so many things that it is vague. Perhaps that was the editor's point. Perhaps "extracted" does not apply to all metalloids. If so, I still think that other, more accurate, more descriptive verbs can be found. I also don't think "can be", "can instead be", "can be" is necessary. I think the more direct "is/are" is better. Metalloids are extracted from ores, metalloids are separated from matrices, etc. Then explain how. In other words, first use the right verb -- an accurate, descriptive verb -- then use is/are rather than can be for the specific methods. And "instead" is not necessary. It's one way, another way, a third way, etc. (It's all right to use "can be + verb" once: "A metalloid can be + verb (extracted/removed/separated etc.) in X ways. One way is to... / One way is by + ing.... Another way to ... is ...." It just doesn't sound good to keep repeating "can be".) That's my opinion, anyway. CorinneSD (talk) 15:06, 14 April 2014 (UTC)
 * Thanks for your thoughtful feedback.


 * I changed “rigorous” to “widely agreed”.


 * Re the change from ‘extracted’ to ‘obtained’. That was my effort at varying phrasing. The sources I was looking at used ‘extracted from’, ‘obtained from’, ‘isolated from’ or ‘prepared from’, so I followed suit.


 * I started with 'can’ rather than ‘is/are’ for the reason that there are multiple ways of extracting a metalloid from an ore or its compounds.


 * There's only one other 'can' now. Does that help? Sandbh (talk) 12:09, 15 April 2014 (UTC)


 * Thank you for your courteous reply. Regarding the definition, it's got to be "widely agreed upon", "widely accepted", or "universally agreed upon", or "universally accepted". I picked "widely agreed upon". It's the way most people speak, and also correct. I appreciate your effort to vary the verbs. Now that there are fewer "can be's", it sounds better. I made a few more edits to improve clarity and conciseness. I hope you approve. (Read the section through as it appears in the article, not just in the Revision History where the highlighted edits can distract one from the way it sounds overall.) CorinneSD (talk) 15:32, 15 April 2014 (UTC)
 * Reads well. I made a couple of refinements to B and Te. Sandbh (talk) 01:27, 16 April 2014 (UTC)


 * O.K. I just have a few questions:


 * 1) Regarding the last edit, with the 1.8 or 1.9 to 2.something, I'm sure it's clearer than it was, but it is still not completely clear. Is it "1.8 or 1.9" because (a) no one is sure, (b) it varies in an unpredictable manner -- sometimes it's one, sometimes the other, or (c) it varies in a predictable manner depending upon something else? The "or", without further explanation, leaves the reader wondering.
 * Electronegativity values are given as either 1.8–2.2 or 1.9–2.2, depending on the author. There are two ranges as there are different ways to calculate EN and no universal :) agreement as to the one true method. The most commonly used method was the one originally proposed by Pauling. Things got complicated because his original values were later revised. Basically, 1.8 to 2.2 works for original Pauling; 1.9–2.2 is better for revised Pauling. For all that, most of the different scales show a high degree of correlation. So both 1.8–2.2 and 1.9–2.2 are found in the literature. Sandbh (talk) 07:11, 17 April 2014 (UTC)


 * 2) Regarding the earlier edit in which you added "following further processing" in parentheses after 97% pure boron, does that "further processing" take place before or after the "processing" described right after that, or is it actually one and the same as the "processing" described right after that? The placement of "following further processing" creates confusion. I think something is needed to clarify when this "further processing" takes place (in relation to the other information there). Also, I notice that in the very next sentence, the steps to creating >99% pure boron are described. I know that, to a scientist familiar with all this, it sounds like I am nitpicking, but to a general reader like myself, these stand out as points of slight confusion. If you need help wording whatever you want to say, just let me know. CorinneSD (talk) 19:52, 16 April 2014 (UTC)
 * The reduction of the trioxide happens first; subsequent processing brings the purity up to 97%, which is about the limit for the trioxide reduction technique. I've edited this passage to try and make it clearer what's going on. None of what you have said is nitpicking. Sandbh (talk) 07:11, 17 April 2014 (UTC)

Spot check of citations
At the request of Sandbh, I have attempted a spot check of a small sample of the citations. I arbitrarily decided to try to check every 60th citation starting with no.27 (chosen randomly) and verify that the information is in the reference. This sample includes 8 citations with a total of 13 references, of which I was able to find 8 either on-line or in my university library; for the other 5 either I have no access or the Google Books excerpts are too short to be useful. I found no instances of close paraphrasing except for very short phrases, so no copyright problems I think.

Here is the detailed list. The numbers 27, 87, ... correspond to the citations as of my two edits of 8 May 2014 (in my time zone).

27 	Claim: P occasionally classed as metalloid
 * Ref.	Thayer			Verified.
 * Ref.	Warren + Geballe (1981)	I have no access to this journal before 1995.
 * Ref.	Masters + Ela		Verified

87	Claim : Copernicium reportedly the only metal known to be gas at rm. temp.
 * Ref.	New Scientist	(information incomplete, add Vol. 67)
 * Prediction based on theoretical calculations, not really verified.
 * Ref.	Soverna		I have no access to this report.
 * Clicking on the link led to the message: Page does not exist or you are not authorized to see it.
 * Link updated---Sandbh (talk) 10:52, 11 May 2014 (UTC)
 * Thanks. I have now verified that this source does verify the claim about copernicium. Dirac66 (talk) 18:23, 11 May 2014 (UTC)
 * Ref.	Eichler		(information incomplete, add Issue May 3, 2007)
 * Experiment shows high volatility, consider as verified.
 * Ref.	Austen	New Scientist	(information incomplete, add Vol. 214). Verified

147	Claim : Sb2O3 = flame retardant
 * Ref.	Carlin		Verified

207	Claim : one synonym used = near metal
 * Ref.	Tyler	 I have no copy available, Google Books excerpts too short to be useful.

267	Claim : Si, Ge, As, Sb, Te solution chem characterized by formation of anions
 * Ref.	Hiller + Herber: I have no copy available, Google Books excerpts too short to be useful.

327	Claim: Sb has extensive organometallic chem
 * Ref.	House: I have no copy available, page not included on Google Books.

387	Claim: Se can be drawn into thin threads when molten
 * Ref.	Regnault 1853, available on Google Books.
 * Verified, though molten and viscous would be more accurate.
 * Done Sandbh (talk) 11:06, 11 May 2014 (UTC)

447	Claim: C, P, Se, I have intermediate electrical conductivity and light sensitivity
 * Ref.	Lutz, verified re metalloid properties from Google Books, list of elements not specified.
 * Page should be 17, not 16. Error may be because p.16 is at end of chapter and skipped in Google :Books, probably a blank page to divide chapters in the paper copy. Dirac66 (talk) 00:59, 10 May 2014 (UTC)
 * Page number corrected---Sandbh (talk) 11:06, 11 May 2014 (UTC)

"Salts"
The use of the word salts to describe some of the compounds in this article- to a "lay" person it may sound as if simple cations were formed. For example
 * carbon- the salts here are all graphite intercalation species. Some are only stable in in the presence of the conc. acid (e.g. HClO4)
 * Wiberg (2001, p. 795) says, 'By cautious oxidation of graphite in conc. sulfuric acid with oxidizing agents...a blue graphite salt [Wiberg's emphasis, not mine] of composition C$+ 24$HSO$– 4$ • 2.4H2SO4 is obtained...when graphite is treated with other strong acids...in the presence of oxidizing agents, other ionic graphic salts are formed (perchlorate, nitrate, hydrogen selenate, hydrogen phosphate).' Sandbh (talk) 12:02, 18 December 2013 (UTC)
 * They're salts Jim, but not as we know them. Wiberg is quite right they are all graphite but they are definitely not the high school chemistry salt- i.e. water soluble- high melting point- ionic lattice - conduct in the melt etc. Greenwood also calls them "salts" (with the quotes meaning they are called salts but don't take it to literally) and includes them in the intercalation section, see p298. The blue sulfate "salt" is, according to NNG, the first stage with an idealised formula (its non-stoichiometric) of C24+ HSO4-.H2SO4, with a structural similarity to C8K. There is also an article on intercalation chemistry in King, Encyclopedia, p 1568. This describes the reaction of graphite and inorganic acid as an intercalation with graphite oxidation, and give an interlayer separation for the stage 1 sulfate as 7.98A. As graphite can also intercalate with graphite reduction- in which case it gains electrons -  for example  with alkali metals - another example of it forming a salt- but this time a carbide. In my view to be wikipedia neutral either both intercalation reaction types (oxidation and reduction) should be mentioned - which supports both metal and non metal arguments or you leave it out.    Axiosaurus (talk) 14:56, 18 December 2013 (UTC)
 * For elements such as C, P, Se there is not discussion about classyfiying them as metals or nonmetals (or metals or metalloids), but as metalloids or nonmetals. Metallicity of Se is often overestimated, but metallicity of C is generally underestimated, of P is very underestimated. H and Rn are even more electropositive than famous Se. I and S are also flawed nonmetals. 83.31.150.242 (talk) 21:26, 18 December 2013 (UTC)


 * antimony- salts?? formulas suggest salt - but are they really ionic? SbPO4  has a layered structure- bonding and is said to be covalent- with a nearly identical electronegativity to P this is expected. Sulfate- again more covalent than ionic; the perchlorate I can only find this abstract(http://www.readcube.com/articles/10.1002/chin.199147026)- the antimony perchlorate compound  reacts with LiClO4 to give LiSb(ClO4)4 or Li2Sb(ClO4)5 the anions are said to be 6 coordinate, points to a more covalent compound. A possible weasel word here is to use "salt" Axiosaurus (talk) 11:23, 18 December 2013 (UTC)
 * I'll check my references for Sb. Sandbh (talk) 12:02, 18 December 2013 (UTC)
 * Cotton & Wilkinson (6th ed., p. 380) say, 'The 5+ ions do not exist, but for trivalent antimony and bismuth cationic behaviour does occur. Bismuth trifluoride seems predominately ionic, and salts such as Sb2(SO4)3 and Bi(NO3)•5H2O, as well as salts of the oxo ions SbO+ and BiO+, exist' and later (p. 396), 'Antimony has some definite cationic chemistry, but only in the trivalent state...Cationic compounds of SbIII are mostly of the "antimonyl" ion, SbO+, although some of the "Sb3+" ion, such as are known. Greenwood & Earnshaw (p. 591) note that,  'The ability to form stable oxoacid salts such as sulfates, nitrates, perchlorates etc., increases in the order As << Sb < Bi. AsIII is insufficiently basic to enable oxoacid salts to be isolated though species such as [As(OH)(HSO4)2] and [As(OH)(HSO4)+ have been postulated in anhydrous H2SO4 solutions of As2O3. In oleum, species such as [As(HSO4)3], [{HSO4)2As}2O] and [{(HSO4)2As}2SO4] may be present. By contrast, Sb2(SO4)3 can be isolated...' IUPAC defines a salt as, 'any compound in which it is possible to identify at least one constituent which is a positive ion or can be classified as electropositive or more electropositive than the other constituents, and at least one constituent which is a negative ion or can be classified as electronegative or more electronegative than the rest of the constitutents.' Perhaps the real point is that antimony is sufficiently basic to be capable of forming "salts" (however defined), particularly oxoacid salts, whereas nonmetals usually aren't Sandbh (talk) 23:57, 18 December 2013 (UTC)
 * Apologies, I am being unreasonable. The "authorities" call them salts, so we are stuck with that and that should be in the in the article. I will put together an article on antimony sulfate as that is one that every one seems to hang the salt label on and is a key point in the classification of metallicity - I will also beef up the graphite intercalation article on these "salts" as greenwood calls them - the issues can then be be handled outside the metalloid article. I love that IUPAC salt definition -which I hadn't seen - so thank you for that- it is so wide ranging that it borders on being meaningless. The gold book defines salt as  "A chemical compound consisting of an assembly of cations and anions. " - which is bit more to my taste. Axiosaurus (talk) 09:15, 19 December 2013 (UTC)

There are also some information about iodine salts (?) Even monoatomic cations, which are typical for metals: Chemistry of Interhalogen Compounds, P. B. Saxena

(http://

books.google.pl/books?id=nvatWdX1ZWcC&pg=PA116&lpg=PA116&dq=%22iodine+phosphate%22&source=bl&ots=qj1QpG6dMW&sig=9mqW4_

Qir6DvkZqGGmUZOILKbnM&hl=pl&sa=X&ei=sUO8UuX8HIfn4gSl_oGIBQ&ved=0CFcQ6AEwBg#v=onepage&q=%22iodine%

20phosphate%22&f=false);


 * [long URL divided into four pieces to avoid screen run off issues] Sandbh (talk) 12:27, 23 May 2014 (UTC)

I.I.T. Chemistry Vol. 1 P Saxena

(http://

books.google.pl/books?id=IN6GriVAJO4C&pg=SA1-PA332&lpg=SA1-PA332&dq=ionisation+energies+iodine&source=bl&ots=4ou1BkIQNp&sig=xk

mZfB9Y5To1FZ-rMNlMylGEK-U&hl=pl&sa=X&ei=LDy8UvXaNOqG4AT

7voGoCw&ved=0CHoQ6AEwCQ#v=onepage&q=iodine&f=false).


 * [ditto] Sandbh (talk) 12:27, 23 May 2014 (UTC)

IPO4, ICl, ICN, I(CH3COO)3 are ionic conductors in molten state according to this book and it looks that they contain I3+ or I+ cations. S, Se, C cations are (rather?) polyatomic.

95.49.102.188 (talk) 15:22, 26 December 2013 (UTC)

Carbon (esp. graphite) vs selenium (esp. grey) (phosphorus is also similar)
Again gray selenium a metalloid but carbon (graphite) a nonmetal... Irritating.

http://pubs.acs.org/doi/pdf/10.1021/ed078p1686

''Selenium in its usual form, on the other hand, has conductivity well within the semiconductor range and its structure is one of infinite spiral chains of selenium atoms with weak interaction of a metallic nature between the chains (1). Its compounds mostly show a nonmetallic nature for the selenium. The element is therefore metallic in some of its properties but a nonmetal in others. This ambiguity defines it as a semimetal even though it does not adjoin the zigzag line.''

Graphite on the graph at that page has more than 100 times higher electrical conductivity than grey Se, but

''The most easily defensible listing of semimetallic elements is Ge, As, Se, Te, Sb, and Bi. If Bi is omitted, as is traditional, then logic requires that Sb also be omitted.''

Because of lack of C, that list is also wrong. Why not graphite? It is also a semimetal. Se also has far much less conductive allotropes. We compare graphite, not semimetalloidal, but nonoconducting diamond and grey selenium, not red selenium. Yellow allotropes of Sb and As (elements generally more metallic than Se and C) are also not metalloidal. Chemically C and Se have very similar metallic properties (such as P).

About weak metallic bond in graphite:

http://przyrbwn.icm.edu.pl/APP/PDF/112/a112z308.pdf

''This comparison leads to the conclusion that the graphene bonding forces are dominantly metallic not van der Waals. Theoretical calculations by Santos and Villagra [13] with the full Kirzhnits correction support this model. As a con- clusion graphite is an anisotropic crystal with atoms covalently bonded in planes (graphenes) with the admixture of metallic bonding. Exceptionally good electri- cal transport is relative with extremely high mobility along planes. Graphenes are bonded by metallic forces with a very small concentration of electrons and relatively low mobility across graphenes. Metallic forces are a small admixture to the covalent forces in planes and they are dominant forces between planes (may be with a small admixture of the van der Waals forces).''

It is also interesting in what conditions electrical conductivity of grey Se was measured. If it was in the light?

Another text about electrical conductivity of Se:

http://pubs.acs.org/doi/pdf/10.1021/j100212a005

Interestingly, Se has also very nonconductive form: Pure, deoxygenated, glassy Se has a conductivity of 10^-16 - 10^-17 ohm^-1 cm^-1.

Interesting text in German:

http://

books.google.pl/books?id=2Oki7r7aUCEC&pg=PA59&lpg=PA59&dq=%22kohlenstoff+ist+halbmetall&source=bl&ots=XsTUbvaQhA&sig=-f5XEjHvKbnFBLyk2pgf1vbm_nk&hl=pl&sa=X&ei=8Hx3Uon3IdKv7AaSzICgCg&ved=

0CDYQ6AEwAg#v=onepage&q=%22kohlenstoff%20ist%20halbmetall&f=false


 * [long URL split into three pieces to avoid screen run off] Sandbh (talk) 12:31, 23 May 2014 (UTC)

Some elements are classified as Halbmetalle (metalloids) and Metametalle (meta-metals). This classification is rather good. Metalloids are (according to the text): B, graphite, Si, black P, As, Sb, Se [red is rather nonmetallic; C, P and Se itself are "semimetalloids], Te. Meta-metals are: Be, Al, Zn, Ga, Cd, In, Sn, Hg, Tl, Pb, Bi.

194.29.130.244 (talk) 11:01, 4 November 2013 (UTC)

Potential Today's Featured Article (PTFA)
I've listed Metalloid as a PTFA for 4 Oct 2014, here. Sandbh (talk) 12:30, 25 May 2014 (UTC)

Template:periodic table (metalloid border)
periodic table (metalloid border) The last sentence seems to have a writing error in it that makes it incomprehensible. Adam Cuerden (talk) 22:57, 19 May 2014 (UTC)


 * The problem seems to be is that the meaning of the pronoun they is not very clear here. I think it should read Although N, Zn and Rn do not appear in the list of metalloid lists ... Dirac66 (talk) 23:49, 19 May 2014 (UTC)


 * Good call Adam. Thank you Dirac66. I replaced the period before the start of what was the last sentence with a semicolon. Does that help? Sandbh (talk) 23:00, 20 May 2014 (UTC)


 * I think it's better. What about this phrase: "the list of metalloid lists"? What does that mean? CorinneSD (talk) 23:43, 20 May 2014 (UTC)
 * That phrase has a wiki link to it? Sandbh (talk) 01:01, 21 May 2014 (UTC)
 * On another point, the elements shaded blue-gray are "Common to rare". Then there are "Very rare" and "Outliers". What is the value of showing a number of elements as "Common to rare"? I know from reading the article that there is disagreement among scientists as to which are metalloids and which are not, but still, for the average reader, how does the designation "Common to rare" contribute to a reader's knowledge? Is there any chance that some could be designated common and some rare, or less common? CorinneSD (talk) 23:48, 20 May 2014 (UTC)
 * The area covered by the common to rare shading area matches the area covered by the more detailed shading given in the periodic table extract at the top of the article. My intention was to avoid duplication hence the single catch all/summary shading for common to rare used here. Sandbh (talk) 10:29, 21 May 2014 (UTC)
 * Subthread about colors & graphics moved from here to below. Is not about actual content.  -DePiep (talk) 06:40, 22 May 2014 (UTC)
 * I think the problem here is more fundamental than the choice of colors or other ways to convey the information. I think the problem is that the information is meaningless. What do common and rare actually mean in this context, and is the information verifiable? Does common for example mean that 50 to 90% (or some other defined percentages) of some sample of references call the element meaningless, and rare perhaps 20% to 50%?? Has anyone actually counted the references? And if so, who chose the sample and what would happen with a different sample? I think we should remove all the semi-quantitative terms (common, rare, very rare, ...) and just say that references disagree with each other and give some specific examples in the format A and B list selenium as a metalloid but X, Y, and Z disagree. Dirac66 (talk) 02:30, 22 May 2014 (UTC)
 * Of course. If the information is removed the color issue is moot. -DePiep (talk) 06:01, 22 May 2014 (UTC)
 * I've just made some bold edits to the template: formats, text-moves, and some ce. No content changes IMO. See the editsummaries and  below. If the points made by  bear fruit, the template will follow (that is to say, I do not prejudice any outcome). -DePiep (talk) 08:15, 22 May 2014 (UTC)
 * Just fallen off my chair and recovered: twin issues with content. A zygose identity problem. -DePiep (talk) 10:00, 22 May 2014 (UTC)

Meaning of common and rare etc
G'day Dirac66. Further to your queries above.

The template uses the label "common to rare" in reference to the table in lede, which shows B, Si, Ge, As, Sb and Te as "commonly recognised as a metalloid"; Po and At as "inconsistently recognised"; Se as "less commonly recognised" and C and Al as "rarely recognised". Hence, "common to rare" encompasses all of the elements appearing in the table in the lede.

Common, less common, rare etc are based on Goldsmith (1982); and Hawkes (2001); plus Mann, Meek and Allen (2000); Gupta (2003); Tregarthen (2003); Kotz, Treichel and Weaver (2009); and Gunn (2014); and the list of metalloids lists which lists 194 metalloid lists drawn from the literature.

Goldsmith; and Hawkes

Goldsmith (1982, p. 526), who surveys the history of the term "metalloid", says, "The elements boron, silicon, arsenic, germanium, antimony, and tellurium have been mentioned most frequently, but polonium and astatine are sometimes included in this list. Beryllium, aluminum, carbon, tin, selenium, and bismuth are mentioned only rarely as metalloids or are classified as metalloids only under special circumstances." Here we see the familiar six of B, Si, As, Ge, Sb and Te; the problematic pair of Po and At; and C and Al as rarely mentioned, plus a few others in the same vicinity.

Hawkes (2001, p. 1686) says, "Lists of 'semimetallic' elements differ. Most include all the elements (other than aluminum and beryllium) that are next to the zigzag line separating metals from nonmetals on the periodic table. Most exclude bismuth and selenium, and most exclude either polonium or astatine or neither but not both." Here we see the familiar six of B, Si, As, Ge, Sb and Te; the inconsistent distribution of Po and At; and the usual exclusion of Bi and Se. Hawkes does not give any details as to which lists of semimetals/metalloids he was referring to. Whatever these were it seems that none included, for example, C or Al, since he does not mention these elements in his article yet, clearly, they are mentioned in some lists of metalloids found in the literature.

The rest of the citations

Mann, Meek & Allen (2000, p. 2783) refer to B, Si, Ge, As, Sb and Te as "the recognized metalloids".

Gupta (2003, p. 4) says that B, Si, Ge, As, Sb and Te, "are often together called as metalloids."

Tregarthen (2003, p. 10) writes that, "Only six elements are commonly classified as semi-metals. These are boron (B), silicon (Si), germanium (Ge), arsenic (As), antimony (Sb) and tellurium (Te)."

Kotz, Treichel & Weaver (2009, pp. 60–62) say, "only six elements are generally classified as metalloids or semimetals"; in their tables they show B, Si, Ge, As, Sb and Te.

Gunn (2014, p. 425) says of metalloids that "these generally include boron, silicon, germanium, arsenic, antimony and tellurium".

List of metalloid lists

The list of metalloid lists supports the claims of the above authors. Thus B, Si, Ge, As, Sb and Te are the most commonly included. Po and At are about 50/50 as per Hawkes, and consistent with the literature disputing their metalloid status. Al and C are rarely classified as metalloids per Goldsmith. Beryllium, tin, selenium, and bismuth are also labelled by Goldsmith as "rarely" mentioned, or classified as metalloids only under special circumstances. Broadly, Goldsmith is still on the mark here although he overlooks P, which Rochow (1966, p. 8), in his "classic" (Mann, Meek & Allen 2000, p. 2783) monograph on the metalloids, said would sometimes be convenient to include in the discussion. In Goldsmith's rare category I have distinguished Se in light of how relatively frequently it appears in the list of metalloid lists, and C and Al since they are immediately next to the metal-nonmetal dividing line and that is the focus of the lede, and metalloids generally. Be and Sn are still rare but in light of how hard it is to find them in metalloid lists I have show them as "very rare", along with P.

I wouldn't support reiterating the above in the metalloid article as it's already essentially covered in the text accompanying the list of metalloid lists.


 * Goldsmith RH 1982, 'Metalloids', Journal of Chemical Education, vol. 59, no. 6, pp. 526–7
 * Gunn G 2014, Critical metals handbook, John Wiley & Sons, Chichester
 * Gupta CK 2003, Chemical metallurgy: principles and practice, Wiley-VCH, Weinheim
 * Hawkes SJ 2001, 'Semimetallicity', Journal of Chemical Education, vol. 78, no. 12, pp. 1686–7
 * Kotz JC, Treichel P & Weaver GC 2009, Chemistry and Chemical Reactivity, 7th ed., Brooks/Cole, Belmont, California
 * Mann JB, Meek TL & Allen LC 2000, 'Configuration Energies of the Main Group Elements', Journal of the American Chemical Society, vol. 122, no. 12, pp. 2780–3
 * Rochow EG 1966, The Metalloids, DC Heath and Company, Boston
 * Tregarthen 2003, Preliminary chemistry, Macmillan Education, Melbourne

Does that help? Sandbh (talk) 05:57, 25 May 2014 (UTC)

Colors and graphics

 * Moved discussion about colors to this subthread. Separate from content issues. -DePiep (talk) 06:40, 22 May 2014 (UTC)


 * My notes: 1. It is not needed to re-describe in text the big border (by the pairs or any other way). The border is visual already, and is *not* dependent on colors (not dependent on contrast). IOW, it is graphically OK. The big border could be darker (more contrast).
 * 2. The colors are too similar (maybe because they are too light). They should be more different (there are only four after all), with good contrast.
 * 3. If a colored cell has a marker (like "" for Common &tc.), the listing in text below is not needed either. (WP:ACCESS). -DePiep (talk) 02:09, 21 May 2014 (UTC)
 * The text re describes the big border in order to meet Wikipedia accessibility guidelines. I don't mind what the colours are but they have to be sufficiently different from the colours used in the periodic table extract at the top of the article. I'm not keen on markers for common etc as we would still need duplicate text for accessibility purposes. Sandbh (talk) 10:29, 21 May 2014 (UTC)
 * The colors have a legend right below, so being different for the PT colors is not that big an issue. About accessability: WP:ACCESS
 * Ensure that color is not the only method used to convey important information. So when you add a code or text to a cell, there is no need to repeat the list below. Simply, because it is shown not by color alone. Even more simple for the fat border: that is not a color effect at all. It is a graphic effect (lines). So as it is now, there is no need to repeat it in a description. It is not even a "by color" thing. (And by the way, that listing in text is undigestable. It should be deleted at any cost). -DePiep (talk) 20:31, 21 May 2014 (UTC)
 * I kind of agree that the text below the table is really not necessary. All this is discussed in the article, isn't it? I don't mind the colors. I think it's better that they are different from the colors in the periodic table. DePiep, is your concern that the colors are too similar so that a person with poor eyesight or color-blindness would not be able to tell the difference between them? Would using patterns such as tiny dots in one section, light cross-hatching in another, tight wavy lines in another, help? (I don't even know if that's possible.) I think adding a letter such as "C" for "Common" in a box would have the potential to confuse readers. The elements are letters, and adding more letters would be confusing. But I recommend removing all, or most of, the text. CorinneSD (talk) 23:38, 21 May 2014 (UTC)
 * Let me refine: the fat border is not "by color only", and so does not need another noting: no repetition in text (it is no "color" at all). I'll add it to the legend, why not.
 * The cell background color meaning (key or legend) now *is* "by color only" and therefor *must* be clarified otherwise (so for now, we must keep the textual description). But when that cell property (the legend color meaning) is added to the cell in writing or in graphic, that listing below can go. All this is straight WP:ACCESS, which is derived from W3C's Web Accessibility Initiative (that is about All Internet Webpages, really).
 * A cell addition could be text, a code letter, of a ordinary footnote (corresponding with the legend). Background hatching or dotting is forbidden, because it makes any letters into a captcha challenge. A better idea is to use cell border graphics (like in this; note that this is irrelevant to colors - good). For this we'd have to separate cells though (whitespace between). -DePiep (talk) 06:01, 22 May 2014 (UTC)
 * Personally, for me the colors are too light to differentiate. I can not link a cell color to a legend box. (I'm not color blind, and my screen is showing quite light. The Outliers look white here). I do not know which other colors "nearby" you refer to. The partial PT in top of the page? Well, there are enough colors to choose from. -DePiep (talk) 06:01, 22 May 2014 (UTC)
 * I've changed the whole appearance of the template, but not wrt the topics discussed here. See editsummary for description of changes. Into a more regular box I aimed. -DePiep (talk) 07:34, 22 May 2014 (UTC)
 * ... but now I did remove that repetition of border elements from the text. Border+legend are enough. -DePiep (talk) 07:54, 22 May 2014 (UTC)
 * Now I (brutally?) moved the listings to the legend, to show up like: . This nicely ties them 1:1 to the legend & colors, and it nicely takes the listings out of running text.
 * From here, we could remove color descriptions out of the text (better not write "the pale blue shadings ..."). We can simply & clearly refer directly to their legend grouping name: "The outliers ..." etcetera. I'll leave it to others (?) to rewrite the remaining text. -DePiep (talk) 08:09, 22 May 2014 (UTC)
 * As it looks good now (to me), I think this setup could be applied to Periodic table (metalloid) too. -DePiep (talk) 09:02, 22 May 2014 (UTC)

DePiep: I owe Dirac66 a response re the meaning of common, rare etc but his comments are harder to address so I'll respond to you first.

The original text that duplicated the content of the tables was put there in response to User:RexxS comments, here, about accessibility, including how well screen readers would be able to read those tables. I suspect that the well-intentioned changes you have made will not be well-understood by screen readers. Certainly, it is not now possible, in the absence of descriptive text, for a blind person to be able to work out where the dividing line between metals and nonmetals is. Sandbh (talk) 06:50, 24 May 2014 (UTC)
 * You are right. The border must be described (as WP:ACCESS#Tables says: not just colors, but all semantic formats must have a description. So I re-added the border description to both templates (see the templates below). My details:
 * I only mentioned the horizontal split, that does enough (exception needed: Li vs. H). I used text like "between Al and Si" because (see MOS:NDASH: "Do not mix en dashes with prepositions like between and from: between 450–500 people."). This fits with the understanding that "between Al and Si" is not a range, but a split. So no "Al–Si" here.
 * I added the list to the legend (not into the extensive text). It leaves thee text for easy reading sentences. That invites for the text to address the border directly, not through a format reference: The metal–nonmetal border is a ... for The grey staircase-shaped line is a .... -DePiep (talk) 07:52, 25 May 2014 (UTC)
 * Question for : about the first/lefthand template: shall I add the member lists to the legend, and cut them out of the text (as done in table #2/righthand)? Or would you like to have more time to consider this? (For now, an easy go/nogo reply is enough for me). -DePiep (talk) 10:00, 25 May 2014 (UTC)
 * I don't mind either way. The key thing for me is what a screen reader would make of the legend boxes. That is why I included the member lists in the text box, as a fail safe. We should ask User:RexxS. Sandbh (talk) 12:25, 25 May 2014 (UTC)
 * Done. If you happen to read this, please take a look at the remaining prose (getting rid of textual color descriptions). -DePiep (talk) 15:51, 25 May 2014 (UTC)
 * You don't need my advice: between yourselves you've addressed the issues and you've agreed that we should make sure as far as possible that screen readers can pick up all the information presented. Cheers --RexxS (talk) 20:46, 25 May 2014 (UTC)
 * You don't need my advice: between yourselves you've addressed the issues and you've agreed that we should make sure as far as possible that screen readers can pick up all the information presented. Cheers --RexxS (talk) 20:46, 25 May 2014 (UTC)

Twin confusing templates
Left: Used in Metalloid. Template /sandbox. Right: Used in Metalloid. Template /sandbox.

Third: Used in Dividing line between metals and nonmetals. Template /sandbox.


 * Added a third template, used in Dividing line between metals and nonmetals: periodic table (dividing line between metals and nonmetals) -DePiep (talk) 08:46, 25 May 2014 (UTC)

Both templates appear in metalloid, and both are based upon List of metalloid lists.

Apart from presentation forms. I understand that the very same List of metalloid lists list is presented here in two ways, each with its own set of descriptions (check how "common" and "rare" is used!). I think their content should be brought to align (say the same when it is the same, do not when not), and somewhere in the topic an explanation must be available on why there are two sets of criteria used. -DePiep (talk) 09:46, 22 May 2014 (UTC)


 * DePiep: Regarding the two different periodic table extracts. The first one is anchored around the dividing line between metals and nonmetals. It is along this line that metalloids are traditionally found, especially B, Si, Ge, As, Sb and Te, but also sometimes either Po or At, and less often (but more frequently in the environmental literature) Se. Since this table is in the lede it is only a preview of what is a more complicated picture.


 * The second table summarises the more complicated picture. The descriptors are based on the same structure as that used on the first table. The first table uses commonly, inconsistently, less commonly and rarely. The second table compresses the entire content of the first table into one gripped up descriptor called 'common to rare', which colours all the elements commonly, inconsistently, less commonly, and rarely recognised as metalloids, the same. After that come the elements very rarely recognised as metalloids, and then the outliers. I designed the table this way so as to not duplicate information already presented in the first table.


 * I see no confusion between the two tables. If confusion remains perhaps some additional explanatory text would help? Sandbh (talk) 06:47, 24 May 2014 (UTC)
 * I have added a third template, in use, that uses this classification. -DePiep (talk) 08:46, 25 May 2014 (UTC)


 * (Note: This is about two templates; the newly added third table could follow the conclusions).
 * (This response did not take into account by Sandbh, above. Also, since I am the one starting a second talk on this subject, this should be of secondary to Diracs66 discussion, above). -DePiep (talk) 09:43, 25 May 2014 (UTC))
 * I see no confusion either. Quite the opposite: I am surprised to discover they are two of the same! Namely: both based on the Loml percentages. I only found this when studying and comparing the two (putting them side by side here helped). Allow me to disagree on your "more complicated" point about the second table. I don't think it is; it has the same setup (the Loml base), only a shifted set of criteria (the percentages). That does not add another layer of complexity I'd say. All clear to me today, but I'm not the target reader.
 * As for explain in the body text: surely needed IMO. I have been working with these a long time, and still did not see the congruence. So our target reader might experience this too. Could it be we need to point more clearly to the "percentage of mentionings in Metalloid lists" base? That would be everywhere then.
 * I find the descriptors not crisp enough, especially since they are spread over two tables while being a single set (almost, #5 is a bit off).
 * Commonly recognised as a metalloid
 * Inconsistently
 * Less commonly
 * Rarely
 * Common to rare [=those four together]
 * Very rare
 * Outliers
 * There are some language hooks ("recognised as metalloid" misses in the second table; so we end up with both adjectives and adverbs; those "rarely" in #4 appear as "rare" in #5).


 * The meaningful wording of descriptors could be better. I get no distinction of intensity between Rarely -- Very rare -- Outliers, to me they have no order. I suggest we try to find better descriptors, keeping in mind the full list while they are being split over two legends. Personally, I have this crude idea: add the percentages? So like (to get the idea):


 * I think we need a single legend table for this either in Loml or in Talk:Loml. (I note that Loml does not explain the colors used). It can have the colors, criteria, and descriptors used. That would be like the grand central legend.


 * I think the tables should converge to a consistent presentation (I'll leave formatting issues out here, this is about content). That is for all pages involved now. I count three templates, four articles. -DePiep (talk) 09:15, 25 May 2014 (UTC)
 * Right. I see what you're saying. Let me think about that. Sandbh (talk) 12:35, 25 May 2014 (UTC)
 * DePiep: I adjusted the right table to bring it more into line with the left table. I think this looks OK now. I did look again at the descriptors. In my view there's a limit as to how far we can go with descriptors not found in the literature. From my response to Dirac66, variations on "common" [commonly; less commonly] and "rare" [rarely; very rarely] seem reasonable; "inconsistently" seems reasonable for the "sometimes" twins (Po, At) and that only leaves "Outlier", but I've explained the intent of that term in the text box accompanying the right table. I also think that "Commonly to rarely" [=those four together, as you noted] in the left table is a reasonable way of linking the left and right tables, without making the right table too duplicative and complicated. I hope this is better. Sandbh (talk) 12:33, 26 May 2014 (UTC)

I'm sure you know what you're doing, but, regarding the left-hand table (with the different colors), which is less common -- "inconsistently" or "less commonly"? And what's the difference between those terms? Sorry to throw this in at this point in the discussion, but, from the point of view of an average reader, really, we have to think about what words mean. CorinneSD (talk) 16:16, 26 May 2014 (UTC)
 * I think the best solution is to replace all these vague words by percentage ranges, as de Piep suggests above. If we say that element A is on 25-50% of (a certain sample of) metalloid lists while element B is on 50%-75%, then any reader over six years old can tell which is on more lists. Someone has to define a sample which can be described in a footnote, and someone has to do the counting. Dirac66 (talk) 19:01, 26 May 2014 (UTC)

I added the median percentages, and kept the words used in the literature (I think this is important), vague as they are. Better? Sandbh (talk) 12:12, 27 May 2014 (UTC)
 * I think the %'s are an improvement indeed. Shouldn't they have some link to an explanation? e.g: (93%)[n 1], with [n 1] a footnote in the template box, saying something like "% mentionings in 194 lists of metalloids" (phrasing...). Footnote can be the same for each % number. Because whatever we are explaining in the two tables, most of it is similar. The "main article:Loml" link better be in that %-footnote, in a sentence. Makes nice reading, and we are supposed to explain/link/mention the Loml anyway. As a hatnote, it formally is not even part of content (but a navigation aid).
 * I don't think the literature descriptors are that important (as Sandbh says). For us it is important that the whole scale makes sense, intuitively even. Terms from literature are used indeed, but more isolated. (For example, "outliers" could be "almost never" or "in special cases": to fit the scale wordings & percentages).
 * I especially don't like "inconsistently" here, because that suggest states that the mentionings in the Loml are inconsistent overall, but of course nobody is claiming that the Loml is consistent. We can only expect consistency within a single list, by a single author. Between lists & authors, every difference can be labeled "inconsistent". IOW: a meaningless word, statistical only.
 * re : to me, your contribution looks in the right place & time. No problem. -DePiep (talk) 17:17, 27 May 2014 (UTC)
 * Oh, and I keep hammering: "Elements with pale blue shading ..." in the prose better be changed into "The outliers ...". No (indirect) color descriptions needed when we have descriptors at hand. -DePiep (talk) 17:20, 27 May 2014 (UTC)
 * To me, the percentage ranges make much more sense than the literary terms. If I may add one more comment: to me, of all those literary terms, "Outliers" makes the least sense. I've hardly ever heard that word used, anywhere, and it's vague. CorinneSD (talk) 17:29, 27 May 2014 (UTC)
 * Much improved now, thank you Sandbh. Adding the percentages defines the terms approximately, and I agree that the median is more significant than the range which could be distorted by a single extreme value. And the link to List of metalloid lists defines the sample and locates more detailed information for the reader. Dirac66 (talk) 18:57, 27 May 2014 (UTC)

Can I address the descriptive terms first. They (common etc; rare etc) are important as they are found in the literature. Using them anchors the article in the literature in accordance with standard WP practice. A term for Po and At is hard. Goldsmith says 'sometimes'. Hawkes effectively says the same. I've used 'inconsistently' to reflect that, compared to the commonly recognised metalloids, Po and At are inconsistently recognised from list to list i.e. about half as often. They are inconsistently recognised due to most authors not doing their homework about the properties of these elements so they guess which element might be reasonable to show as a metalloid or uncritically copy another author's uncritical or guesswork classification. Hawkes also published articles disputing the classification of Po and At as metalloids, as did Holt, Rinehart and Wilson (ref 22). (Given these circumstances it is no surprise that the median appearance frequency involved is about half that of the commonly recognised metalloids.) The disputation was mentioned in the text box. Should I add it again? Or is there a better term?

'Outlier', AFAIK is a fairly ordinary word, meaning an observation point that is distant from other observations; something that lies outside the main body or group that it is a part of, as a cow far from the rest of the herd, or a distant island belonging to a cluster of islands; a value that lies outside of other values etc. Personally, this term seems quite appropriate. And the text box explains the outlier nature of the elements in question. I'm open to suggestions for a better 'neutral' term. 'Almost never' is right in spirit but not objective enough nor neutral enough; 'in special cases' isn't accurate as there is nothing intrinsically special enough going on. (this comment is part of Sandbh 12:06 edit )


 * Well, maybe the problem for me is that most of the terms refer to time (frequency) -- commonly referred to..., occasionally, less common/less commonly, almost never, never, inconsistently -- but "outlier" or "outliers" refers to space. I think that, to be consistent, they should all refer to time/frequency or all refer to space. The example of the cow far from the other cows and the island far from the other islands make sense as illustrations for "outlier". Another use, which I have heard, is to describe a house that is far from other houses, far from the village. Since "outlier" is the only one that refers to space, it should probably be changed to one that refers to time/frequency, such as "rare" or "rarely", or "infrequent" or "infrequently". The question is, do you wish to describe the frequency with which the element is mentioned by the various scientists in the lists of metalloids or the physical location of a percentage on a graphic representation of all the percentages? "Outlier" can clearly be used for the second. Maybe the word can also be used for the first, transferring the meaning from an outlier in space to an outlier in relative frequency. I guess that's what you're suggesting. If you're all comfortable with that, I certainly won't object. CorinneSD (talk) 15:32, 28 May 2014 (UTC)
 * Yes, it was the outlier in space to the outlier in frequency meaning that I had in mind. The outlier article captures this meaning, and offers some possible explanations for the occurrence of outliers, some of which are quite good in the context of this discussion. I haven' stopped thinking about alternatives to 'outlier' and so far haven't been able to identify anything better for what is a tricky concept. Sandbh (talk) 22:52, 29 May 2014 (UTC)
 * I skimmed the article on outlier. I guess the reason I had not often heard the word except in literature is because I never studied statistics. I guess the word makes sense. I thought of a few possible alternatives, but you may decide that outlier is still the best word: erratics, eccentrics, peculiars (not really a word), isolates, and exotics. CorinneSD (talk) 02:05, 30 May 2014 (UTC)
 * I changed "Outlier (–)" to "Sporadically". According to the Oxford English Dictionary, 2nd ed., meanings of sporadically include: "Scattered or dispersed, occurring singly or in very small numbers, in respect of locality or local distribution; of single persons or things: accidental; isolated; Astr. applied to a meteor that is isolated and does not appear to belong to a shower." I bolded the more relevant parts. Better? Sandbh (talk) 13:15, 30 May 2014 (UTC)
 * Sporadically is fine, and as an adverb it parallels the other words. I had been looking for a noun since outlier(s) is a noun. CorinneSD (talk) 15:34, 30 May 2014 (UTC)
 * Excellent, thank you CorinneSD. Sandbh (talk) 01:28, 31 May 2014 (UTC)

Response to the rest of DePiep's comments later. Sandbh (talk) 12:06, 28 May 2014 (UTC)
 * DePiep: I try and avoid notes next to numbers where I can because I think such a combination can look cluttered. You also seem to be suggesting adding notes to all of the percentage ranges which I think would be overkill. If the text in the text box accompanying a table is supposed to elaborate what the table shows then I think the current layout explains what the percentage ranges are well enough and that the links to the LoML are reasonably evident.


 * On 'inconsistently', your point is well made about different authors. In this case the main descriptors used by the two authors are common and rare, which, when expanded into commonly, less commonly, rarely, and very rarely seem reasonably hierarchical to me. With the addition of 'inconsistently' in between commonly and less commonly, and the percentage ranges, I think this works well enough. Thus, comparing a median of 93% and 44%, the latter seems inconsistent in terms of how many metalloid lists I would see with Po and At, compared to how many I would see with the recognised metalloids i.e. nearly all. And then (even) less commonly at 24% is, indeed, less common than 44%. So, I think inconsistently is still OK, given how hard it is to reasonably capture what the literature says. Sandbh (talk) 22:40, 29 May 2014 (UTC)
 * re the numbers (ref links & footnotes): remove & prevent them as you like. But we must provide a concise explanation for the core of the template: the difference in metalloid-ness, by authors. That is the first and foremost task. The secondary task is to describe the metal-nonmetal dividing border. For this, if and when a footnote/ref is needed, then we use it. If a wikilink is needed, we use it (I expect the Loml link here). (Technically & nicely, we can add regular refs to a template; I'll show demo's if you ask).
 * re the descriptor "inconsistently": you have described here where it comes from, but that does not address my point: "inconsistence" states a flaw in logic, both in science as in common language. If a single author, present in the Loml, would be inconsistent, that likely would be addressed in literature ("author X does not apply their own logic consistently"). But for the whole list, there is no general logic to check consistency against. (Check this: by your description, each element that does score 100% is to be labeled a metalloid inconsistly). But there is no logic to check the lists against, there is no superauthor that wrote the whole Loml. It is just a statistic overview. The descriptor "inconsistently" better be something like "irregularly", "sometimes it is, sometimes not", "variously": everything that takes the faulty logic claim out, and leaves the statistics claim in. -DePiep (talk) 06:25, 30 May 2014 (UTC)
 * "Irregularly" is pretty good, and I've changed the template on this page to that, thank you DePiep. Sandbh (talk) 12:52, 30 May 2014 (UTC)


 * About the germanium detail. Recently, a germmanium detail was added in the prose below. I think this has no place in this template. First it is not marked in the graph as such, so there is nothing in the prose to be explained at all in the first place. Second, it is a single-element detail, of which more could exist (why not added more? -- but I do not want to promote adding more). Thirdly, it is about the metal-nonmetal border, not the metalloid-ness (that is secondary to the whole "metalloid" topic, right?). I expect all such details be in the article body text if needed, but not squeezed in as a sidenote in the table. Fourthly, it uses the wording "on the wrong side", which makes a reader like me wonder: "where is that good/wrong judgement made? Can't find it". The better wording is already there: "arbitrary" says it all, and neutrally. In short: rm the prose about germanium completely, and check the article for having this in its body. -DePiep (talk) 06:44, 30 May 2014 (UTC)
 * Ge content removed; already alluded to in main body of article. Sandbh (talk) 01:53, 31 May 2014 (UTC)


 * In the third table, I have added aluminium to the list (colored bg & linked). Slavishly following the two metalloid templates. -DePiep (talk) 06:46, 30 May 2014 (UTC)
 * Oh. No. That table only shows elements commonly or inconsistently irregularly recognised as metalloids, not Al which is rarely recognised as such. Sandbh (talk) 12:29, 30 May 2014 (UTC)
 * After studying this, I understand this is another variant schnitt of the same. I reverted. This is making me sick. I pity the more casual reader. -DePiep (talk) 01:47, 31 May 2014 (UTC)


 * About other issues, not content (MOS, guidelines, readability). I have started two sandbox version copies. Proposing non-content changes there: see,.
 * 1. Symbol lists per legend entry: regular font, not small. They are in the sentence now, not bracketed any more.
 * 2. Removed the "Main article: Loml" note in the left one. Instead, wikilinked that page in the prose text, in both tables. Explained: a "Main" link is a hatnote and non-content always by guideline. But in the cases here, it is part of the description. A regular wikilink serves all very well. Also reduces cluttering of the whole, as a bonus.
 * Copy them into live versions? -DePiep (talk) 06:53, 30 May 2014 (UTC)
 * Yes please. More to follow from me, too. Sandbh (talk) 12:11, 30 May 2014 (UTC)
 * OK, I made those changes; Ge still to be addressed. Sandbh (talk) 01:44, 31 May 2014 (UTC)
 * I think we may be done now? Sandbh (talk) 01:53, 31 May 2014 (UTC)
 * ... Almost! See my remarks oon Vernon below. I've reread this whole thread & more, and the result looks very well. Keeping in the terms from literature in turns out strong too. (Now it occurs to me: could we move the page name of List of metalloid lists into something more self-explaining? Mention its scientific base somehow? As it is now, it is both correct and cryptic). My remaining notes on the graphics go in another subsection. -DePiep (talk) 08:16, 2 June 2014 (UTC)
 * The name of "List of metalloid lists" came from an article by Hawkes called "Semimetallicity". He starts off by saying "Lists of "semimetallic" elements differ." Hence the list in question is a list of metalloid lists based on a sampling of metalloid lists found in the literature. Did you have anything mind for an better name? Sandbh (talk) 07:21, 5 June 2014 (UTC)


 * Let's take a look at the Vernon (2013) reference. First: what is its substance & importance? How does it determine the List of metalloid lists and its application here? There is no quote or statement at all. Then, when the Vernon ref substance were added, is it that important that is should be mentioned here, singled out? Are there no other references in the Loml that are nearly determining? Maybe move this one into the Loml page with the other ones, where the list can be refined more refined, or is it enough being in the article body text? Second: if it stays, can we turn it into a direct reference (not through a group footnote)? (Technically, one can put the named ref definition " in both pages template and article, and it will show up as expected. Be sure to have them as exact copies). -DePiep (talk) 08:16, 2 June 2014 (UTC)
 * I have an open mind about the Vernon ref. It's there to anchor the list in the literature, as the article (Vernon 2013) refers to the list of metalloid lists. I suppose you could turn it into a direct ref rather than a note. It will be a little odd since Vernon (2013) actually cites the list of metalloid lists rather than duplicating it in itself. What do you think? Sandbh (talk) 07:32, 5 June 2014 (UTC)
 * While we are talking footnotes: could you reconsider the [n 1] link that is now in the title of the righthand template? This one has substance, but to me does not look defining enough to be in the title. At least it better be in below in the text, and maybe even better it belongs in the Loml page. The template is not suitable for details (the article body text and Loml page are). Or, arriving from the opposite direction: why does it have to be in there? -DePiep (talk) 19:27, 2 June 2014 (UTC)
 * Removed as it's redundant, thank you DePiep. Sandbh (talk) 07:12, 5 June 2014 (UTC)


 * I have added (copied) into the the legend for the border. Now the text is a bit off: it has "the border" vs. "arbitrary". Can you take a look & an edit? -DePiep (talk) 08:28, 2 June 2014 (UTC)
 * btw, I see you removed the "between Uus and Uuo" description . Intentional? -DePiep (talk) 08:33, 2 June 2014 (UTC)
 * OK, will have a look at all of these although I may not be able to do so until later in the week. Sandbh (talk) 07:28, 3 June 2014 (UTC)
 * Changed the reference to "the" metal-nometal dividing line to "a typical" etc. Uus and Uuo text ref restored; I probably unintentionally removed it during some copy and pasting.
 * Added hatnote to say that the "confusion" in section title is resolved. Templates now build from the same set-up and text base. -DePiep (talk) 11:57, 16 June 2014 (UTC)

June 15, 2014 proposals
These twin tables have evolved nicely since my last visit to metaliods, but some issues remain. The category "commonly to rarely", spanning from 9% to 99% of cited lists, is much too broad. The label is confusing and this grouping obscures, rather than highlights, trends. The "sporadically" label includes several elements with separate citations (N, Zn, Rn), but not those with just a single citation in the loml (i.e. H, Ga, Pb, Fl, Uus, Uup).

My recent edit was reverted, so i will lay out my reasoning here. The label "commonly" for inclusion in 86% to 99% of cited lists is simple and distinct. I grouped Se (24%) with Po (48%) and At (40%) as "irregularly". I also separated Be (4%), C (8%), Al (9%), P (5%), Sn (3%), and Bi (6%) to a label "occasionally". This Leaves S, I and Lv, each with 3 listings (1.5%), which i included with those having fewer mentions and labeled "rarely". Finally, I changed "rarely" in the lead table (C and Al) to "occasionally", for consistency. I think that this grouping is is clear, provides more detail, and helps to show the gradual trend from metals to nonmetals. Bcharles (talk) 15:08, 15 June 2014 (UTC)
 * I did that reversal, bold & AGF,, because I remembered it was discussed here. I'd like to see checked that the changes are in line with earlier contemplations in thes thread (that also includes lower order issues like coloring). For example, I remember that the % mentioned is a median, not a range. And that the descriptors (in the legend) are derived from literature. And that &mdash;a very fine point&mdash; the descriptors are spatial, not temporal or statistical. I have no judgement on the change: it could contain improvements. -DePiep (talk) 16:16, 15 June 2014 (UTC)


 * I don't think that i broke anything discussed under coloring. The median percentages are used in the lead table, but  in the 'metaloid border' table only the first category mentions percentages, giving a range of medians(?!).  The term 'commonly recognised' seems to be widely used and clear. The only other term that is cited as coming from the literature is rarely, but it is unclear to me how many sources use the term and what elements they indicate by it. In a survey of a dozen sources, one mention could be described as rarely.  However, in a survey of 200 sources, 9% is not rare.


 * I am not stuck on any particular terms but a more specific grouping is called for in the 'metaloid border' template. Bcharles (talk) 18:18, 15 June 2014 (UTC)
 * Noted all the above; let me ponder for a bit. Sandbh (talk) 07:52, 16 June 2014 (UTC)

Formatting the definitions section
I've come up with three options for formatting this section in such a way as to reduce the impact of or eliminate the need for embedded lists:

Option 1 The generic definition set out at the start of this article is based on metalloid attributes consistently cited in the literature. Illustrative definitions and extracts include: (a) 'In chemistry a metalloid is an element with properties intermediate between those of metals and nonmetals.'&lt;ref>Cusack 1987, p. 360&lt;/ref> (b) 'Between the metals and nonmetals in the periodic table we find elements…[that] share some of the characteristic properties of both the metals and nonmetals, making it difficult to place them in either of these two main categories.'&lt;ref>Kelter, Mosher & Scott 2009, p. 268&lt;/ref> (c) 'Chemists sometimes use the name metalloid…for these elements which are difficult to classify one way or the other.'&lt;ref name="Hill 2000, p. 41">Hill & Holman 2000, p. 41&lt;/ref> (d) 'Because the traits distinguishing metals and nonmetals are qualitative in nature, some elements do not fall unambiguously in either category. These elements…are called metalloids…'.&lt;ref>King 1979, p. 13&lt;/ref> More broadly, metalloids have been referred to as: (e) 'elements that…are somewhat of a cross between metals and nonmetals'&lt;ref>Moore 2011, p. 81&lt;/ref> or (f) 'weird in-between elements.'&lt;ref>Gray 2010&lt;/ref>

Option 2 The generic definition set out at the start of this article is based on metalloid attributes consistently cited in the literature. The following are some illustrative definitions and extracts from different authors:

Intermediate nature 'In chemistry a metalloid is an element with properties intermediate between those of metals and nonmetals.'&lt;ref>Cusack 1987, p. 360&lt;/ref>

Mixed properties 'Between the metals and nonmetals in the periodic table we find elements…[that] share some of the characteristic properties of both the metals and nonmetals, making it difficult to place them in either of these two main categories.'&lt;ref>Kelter, Mosher & Scott 2009, p. 268&lt;/ref>

Difficult to classify… 'Chemists sometimes use the name metalloid…for these elements which are difficult to classify one way or the other.'&lt;ref name="Hill 2000, p. 41">Hill & Holman 2000, p. 41&lt;/ref>

…unambiguously 'Because the traits distinguishing metals and nonmetals are qualitative in nature, some elements do not fall unambiguously in either category. These elements…are called metalloids…'.&lt;ref>King 1979, p. 13&lt;/ref>

More broadly, metalloids have been referred to as 'elements that…are somewhat of a cross between metals and nonmetals'&lt;ref>Moore 2011, p. 81&lt;/ref> or 'weird in-between elements.'&lt;ref>Gray 2010&lt;/ref>

Option 3 The generic definition set out at the start of this article is based on metalloid attributes consistently cited in the literature. Some illustrative definitions and extracts follow. 'In chemistry a metalloid is an element with properties intermediate between those of metals and nonmetals.'&lt;ref>Cusack 1987, p. 360&lt;/ref> Such elements, which lie between the metals and the nonmetals in periodic table terms, 'share some of the characteristic properties of both the metals and nonmetals, making it difficult to place them in either of these two main categories.'&lt;ref>Kelter, Mosher & Scott 2009, p. 268&lt;/ref> This difficulty in classification arises due to the qualitative nature of the traits used to distinguish metals and nonmetals. The result is that 'some elements [which] do not fall unambiguously in either category…are called metalloids…'.&lt;ref>King 1979, p. 13&lt;/ref> Still other authors take a broader or more whimsical view of the nature of metalloids, referring to them as 'elements that…are somewhat of a cross between metals and nonmetals'&lt;ref>Moore 2011, p. 81&lt;/ref> or 'weird in-between elements.'&lt;ref>Gray 2010&lt;/ref>
 * signing: added by, 28 March 2013 (sign added -DePiep (talk) 08:59, 17 October 2014 (UTC))

boron chemistry
Congratulations on achieving FA status. I know how hard it is to get there, so well done.

I don't have time to make edits as I'm away all next week. In the meantime I'd just like to point out a weakness in the chemistry of boron. There is no mention of the fact that simple compounds of boron are Lewis acids. Boron does not form ionic compounds, by which I mean compounds containing B3+, but there are plenty of ionic species such as borohydride, BH4- which can be viewed as adducts with Lewis bases. Another example: BCl3 + Cl- → BCl4- etc. In the same way, many compounds of group 15 elements like As and Sb can act as Lewis acids, e.g. SbCl5 + Cl- → SbCl6-. I hope to attend to this when I'm back. Petergans (talk) 12:54, 4 October 2014 (UTC)
 * Thank you Petergans. I'll have a closer look at this and look forward to your further input. Sandbh (talk) 03:19, 5 October 2014 (UTC)
 * I mentioned this thread at Talk:Boron. -DePiep (talk) 10:18, 5 October 2014 (UTC)

I have added a rewrite of the whole section on boron. At present the re-write is invisible, being enclosed in I had to do this as I could not be sure of handling the references correctly. Also I have added some personal comments, indicated by (PG), which will need to be removed. Please have a look and eventually replace the current text as you think fit. Petergans (talk) 08:43, 17 October 2014 (UTC)
 * Thank you; will do. Sandbh (talk) 11:24, 17 October 2014 (UTC)
 * Perhaps this high quality summary of B chemistry would be more be more appropriate at Boron (which is weak on chemistry) or as the basis of a new article, boron chemistry and compounds. A stripped down version of this could then be used here.Axiosaurus (talk) 11:45, 17 October 2014 (UTC)
 * I agree that boron is very weak on chemistry and that a new article on boron chemistry would be welcome. However I would not want to do it on my own. I suggest using Greenwood and Earnshaw as the principal source. Norman (I refer to my former head of department) took a life-long interest in boron chemistry going back to his post-doc days with Emeleus. The book reflects his expertise and knowledge the subject. Carboranes, for example, was an productive research area at Leeds. I was not personally involved in it, but obviously knew something of what was going on. Petergans (talk) 15:33, 17 October 2014 (UTC)
 * Have made a start on reviewing your rewrite (i.e. I pasted it into my sandbox). Sandbh (talk) 02:08, 20 October 2014 (UTC)

Petergans, thank you again for your thoughtful rewrite, here.

You proposed changing mention of how "close" boron was to the metal-nonmetal borderline, to how "far" it was. The citation in question uses the adjective "close". I changed the article anyway to refer to the "proximity" of boron.

The article says, "…nearly all compounds of boron are covalent, with a few complexed anionic and cationic species" and supports this statement with two citations. You proposed changing this to say, "All compounds of boron(III) are covalent". Since, for example, Köster (1977, p. 170) writes that, "Although the majority of ionic boron-containing species are anions (see 11.2), there are now many examples of cationic boron complexes, the so-called boronium salts", I haven't actioned this suggestion.


 * Köster R 1977, 'Salt-like boron compounds', in HJ Albert, H Zimmer & K Niedenzu, Methodicum Chimicum: A critical survey of proven methods and their application in chemistry, natural science, and medicine, Volume 7: Main group elements and their compounds, Part A: Main Group 0 to IV, Academic Press, New York

The description of the bonding in boron has been corrected, thank you.

Mention of simple compounds of boron as Lewis acids has been added to the chemical properties paragraph. Ditto.

On Fe(CO)4 I've added a note saying that in the applicable reaction, this is a short-lived intermediate (as is also the case with BH3). I trust this addresses your observation.
 * No it does not. The properties of these species are, I assume, not observed properties but are calculated properties. BH3 may have been observed by matrix isolation at 4K . There is some literature on the BH3- radical anion. Petergans (talk) 18:10, 4 November 2014 (UTC)
 * Fe(CO)4 can be prepared in an Ar matrix at 20 K. See here or here. Sandbh (talk) 06:51, 9 November 2014 (UTC)

I kept the content about borate chemistry, B2O3, and the organometallic chemistry of boron as indicators of metalloid behaviour.

I'm still thinking about whether or not to expand the content dealing with the aqueous chemistry of boron. Sandbh (talk) 09:22, 28 October 2014 (UTC)

The article says, "…nearly all compounds of boron are covalent, with a few complexed anionic and cationic species" This statement is very misleading. All chemical compounds of boron are covalent compounds, including those like BPh4+ Cl- and Na+ BCl4- which are salts of ions containing boron covalently bonded to other atoms. The only compounds of boron that are not covalent compounds are the borides, MxBy. Borides don't contain boron covalently bonded to another element. They are described a intersitial compounds (c.f. carbon steel). A cation such as [B(H2O)4]3+ does not exist; in the context of this article that is important as it is characteristic of a non-metal whereas metals in low oxidation states do form aqua-ions.Petergans (talk) 18:10, 4 November 2014 (UTC)
 * Have updated mention of the bonding in boron compounds (all covalent bar borides). Sandbh (talk) 03:42, 9 November 2014 (UTC)
 * I intend to add some content to all of the mini-bio's of the recognized metalloids, re their solution chemistry. B and Si (nil cationic) are straightforward. Sb and Te (some cationic) are reasonably straightforward too. What is holding me up at the moment (as usual when working in the twilight zone) are the two in between metalloids i.e. Ge and As. Definitively nailing down their chemistry in solution is proving to be quite difficult. Sandbh (talk) 11:09, 20 November 2014 (UTC)
 * Right then. All done. Sandbh (talk) 06:12, 13 December 2014 (UTC)

Germanates, arsenates
There is a good discussion of germanates in Greenwood & Earnshaw, p 382. The same authors (p387) state "Oxoacids of Ge are usually unstable, generally uninteresting and commercially unimportant", but some details are given. Arsenites of alkali metals ,p575, are "very soluble in water". For As(V), arsenic acid H3AsO4 "can be obtained in aqueous solution" and some comparison with phosphoric acid etc. follows. See also M.T. Pope "Heteropoly and Isopoly Oxometalates", Springer Verlag, New York, (1983).Petergans (talk) 10:11, 29 November 2014 (UTC)

Holleman and wiberg states that there are no Ge cations known. p 896 English translation 2001. Same is repeated in latest German version. p 1007 Axiosaurus (talk) 16:21, 29 November 2014 (UTC)
 * Yes, thank you. I was aware of what Wiberg (2001) said. It's odd that this is repeated in the latest edition. A cryptand compound containing a Ge2+ cation was synthesized in 2008. Before that, the layer compound/phase Nb3GexTe6 (x ≃ 0.9) was prepared in 1992, and reported to contain GeII cations. Pan, Fu & Huang (1964, p. 182) presume the formation of the simple Ge++ ion when Ge(OH)2 is dissolved in a perchloric acid solution, on the basis that, "ClO4− has little tendency to enter complex formation with a cation". On this point, Wiberg (2001, p. 899) notes that GeO and Ge(OH)2 are amphoteric and dissolve in both strong acids and bases, forming germanium(II) salts (e.g. Ge(OH)2 + 2HClO4 → Ge(ClO4)2 + 2H2O) although in the same sentence he also refers to GeCl2 as a "salt". As far as I know, GeCl2 doesn't feature a Ge2+ cation. Rich (2007, Inorganic reactions in water, p. 342) similarly says that "perchloric acid dissolves Ge(OH)2 as Ge2+ " but doesn't give a source. I have a few more sources and am waiting for an inter-library loan before trying to make sense of it all. Sandbh (talk) 06:16, 30 November 2014 (UTC)


 * The situation with aqueous Ge2+ is as I undertand it fairly clear. Aqueous solutions of Ge2+ are unstable, the porbaix redox info is in Schweitzer G.K., Pesterfield L.L. The Aqueous Chemistry of the Elements (OUP, 2010). However while unstable thermodynamically in water it does persist for a while. In the presence of complexing agents such as Cl- the GeII state is stabilised and persists (according to this 2000 paper by Babich et al (10.1039/b000401o), for weeks, but obviously as a complex. Wiberg I think is referring to stable "salts" that contain Ge2+ have not been isolated from aqueous solution. Looking at Wells for the germanium sulfate, perchlorate etc. (with various ox. states) he states that these  demonstate Ge metallic character, the key words are "no silicon analogues". However the reference he gives is to a 1963 paper in Acta Scandinavica, (Ingri, Nils,Equilibrium Studies of Polyanions. 12. Polygermanates in Na(Cl) Medium, Pages: 597-617, doi:  10.3891/acta.chem.scand.17-0597) you will be able to download it, Acta C. S. has no paywall. This paper mentions these compounds and gives even older references for them. For me this is a rare weak point in Wells' brilliant book. Axiosaurus (talk) 14:32, 19 December 2014 (UTC)
 * I'm not so sure about Ingri, given he says (p. 599) "Ge forms a rather stable bivalent ion, Ge2+". Perhaps he was referring to the older papers in the RSC journals discussing germanous chemistry. I think the situation re Ge cations is fairly well set out now, in the note attached to the paragraph on Ge chemistry. Sandbh (talk) 10:01, 23 December 2014 (UTC)

Metalloidal nonmetals - C, P, Se
I think that aluminium is too clearly a metal to suggest that it is a metalloid. We can't say too much about astatine because of its instability (it could have metallic properties despite being in one group of periodic table with fluorine, chlorine, bromine and iodine).

I think that extremely high sublimation point (which is obviously higher than melting and boiling points of all nonmentals, including P and Se) of carbon should be mentioned in section about carbon in the article. Carbon is more like boron and silicon than like nitrogen or chlorine (the same is true for P and Se), its high sublimation point and formation of covalent network solid which remains solid in very high temperatures and coducts heat and ever electricity relatively well makes carbon more similar to metalloids than to nonmetals.

For me selenium (considered (in my opinion not properly)) as a metalloid (while P and C are just nonmetals) in German Wikipedia is not more metallic than carbon and phosphorus. For me most metalloid-like elements are: carbon, phosphorus and selenium (which is on one group of metallicity with C and P, not with arsenic or tellurium!).

Black phosphrous has lower band gap in bulk than boron, silicon and germanium. It has also larger thermal conductivity than tellurium. Melting point of other than white allotropes of P are higher than boiling point of white allotrope of this element and of S and Se. Phosphorus pentoxide can exist not only in dimeric form, but also in polymeric, which have really high melting point for an oxide of a nonmetal, not a metalloid (562°C). More stable O' form of phosphorus pentoxide has relatively large density of 3,5 g/m3. — Preceding unsigned comment added by Krobon (talk • contribs) 17:34, 21 October 2016 (UTC)

boron ligands replacing toxic phosphorus ligands?
I have two issues with this statement:

While the smaller alkylphosphines are toxic, arylphosphines, and alkylphosphines of higher molecular weight are not known to be so.

Second, both boryl ligands and Z-type borane ligands have very different electronic properties compared to phosphine or carbene ligands, and there is no indication that they can serve as a replacement in *most* metal catalyzed reactions (e.g. cross coupling, metathesis, etc., etc).

I think this statement should be changed to something like, "boron-centered ligands have been explored in metal-catalyzed reactions."

Alsosaid1987 (talk) 15:07, 14 September 2017 (UTC)


 * The first supporting citation says

Bertrand’s lab at UC Riverside specializes on catalysts. A catalyst is a substance – usually a metal to which ions or compounds are bound – that facilitates or allows a chemical reaction, but is neither consumed nor altered by the reaction itself. Crucial to the reaction’s success, a catalyst is like the car engine enabling an uphill drive. While only about 30 metals are used to form catalysts, the binding ions or molecules, called ligands, can number in the millions, allowing for numerous catalysts. Currently, the majority of these ligands are nitrogen- or phosphorus-based.

The trouble with using phosphorus-based catalysts is that phosphorus is toxic and it can [italics added] contaminate the end products,” Bertrand said. “Our work shows that it is now possible to replace phosphorus ligands in catalysts with boron ligands. And boron is not toxic. Catalysis research has advanced in small, incremental steps since the first catalytic reaction took place in 1902 in France. Our work is a quantum leap in catalysis research because a vast family of new catalysts can now be added to the mix. What kind of reactions these new boron-based catalysts are capable of facilitating is as yet unknown. What is known, though, is that they are potentially numerous.


 * Does this help? Sandbh (talk) 06:06, 17 September 2017 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified 3 external links on Metalloid. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20130203033834/http://cot.food.gov.uk/pdfs/vitmin2003.pdf to http://cot.food.gov.uk/pdfs/vitmin2003.pdf
 * Added archive https://web.archive.org/web/20120426082753/http://zernike.eldoc.ub.rug.nl/FILES/root/2007/JAmChemSocvSetten/2007JAmChemSocvSetten.pdf to http://zernike.eldoc.ub.rug.nl/FILES/root/2007/JAmChemSocvSetten/2007JAmChemSocvSetten.pdf
 * Added archive http://arquivo.pt/wayback/20160516135811/http://research.wpi-aimr.tohoku.ac.jp/eng/research/676 to http://research.wpi-aimr.tohoku.ac.jp/eng/research/676

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 17:44, 24 September 2017 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on Metalloid. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20160817165930/http://www.nist.gov/data/PDFfiles/jpcrd260.pdf to https://www.nist.gov/data/PDFfiles/jpcrd260.pdf

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 09:42, 7 October 2017 (UTC)

Black phosphorus
https://en.m.wikipedia.org/wiki/Allotropes_of_phosphorus#Black_phosphorus

So it depends... I suggest NOT to talk of is white phosphorus a metalloid but "black phosphorous IS a mettalloid" and we just diagonal cut the cell of phosphorus: HAS a metallic configuration. Same, Carbon or Diamond. Or black coal. Divide tbe cell diagonally. And say Yes Grapphite sthsth IS a metalloid. DIA. OND, NO.

ETC Wikistallion (talk) 13:02, 27 October 2018 (UTC)


 * Metalloids (B, Si, Ge, As, Sb, Te) can be distinguished by the weak strength of their oxyacids. The oxyacids of the remaining nonmetals having a metallic appearance (C, P, Se, I) are four to nearly ten orders of magnitude more acidic than those of the metalloids. So, P and C are not ordinarily regarded as metalloids. Sandbh (talk) 18:58, 28 October 2018 (UTC)

Iodine. Why Metalloide?
http://images-of-elements.com/iodine.php Tell me what it looks like in this config

Well Carbides or phosphides or Borides are not always metals. So thats ONE way to determine a metalloide. Are there any metalloide iodides?

Wikistallion (talk) 13:43, 27 October 2018 (UTC)
 * The nonmetals adjacent to the metalloids show some metalloidal character so it is not surprising that there are such things as metallic “-ides”. The existence of such compounds does not by itself warrant classifying the element in question as a metalloid. I don’t know about metallic iodides. Possibly some iodides of transition metals in lower oxidation states could be metallic. Even if they were it wouldn’t count for much. Sandbh (talk) 09:25, 29 October 2018 (UTC)

Ununhexium
I've found nothing to suggest that ununhexium is a metalloid, am I missing something? Ctachme 17:35, 13 Aug 2004 (UTC)
 * Neither is ununseptium, according to its article. I've commented out both. Radagast 19:04, Feb 16, 2005 (UTC)

Conduction and valence bands
OK, I'll admit I'm a new editor, so I'm not quite sure how things work around here. Is it really true that there is no way of telling a metalloid from a semiconductor from a metal? What I've always thought was that metals have overlapping conduction and valence bands, semiconductors have a large bandgap and metalloids have a bandgap small with respect to kT. "Large" and "small" are fairly subjective terms, of course, but I've never heard anyone refer to silicon or germanium as anything other than a semiconductor, or bismuth as anything other than a semimetal.
 * Eric 22:28, 29 Jan 2005 (UTC)

Move discussion in progress
There is a move discussion in progress on Talk:Alkali metal which affects this page. Please participate on that page and not in this talk page section. Thank you. —RMCD bot 06:15, 28 February 2019 (UTC)